---
dataset_info:
  features:
  - name: id
    dtype: int64
  - name: task
    dtype: string
  - name: env_seed
    dtype: int64
  - name: image
    dtype: image
  - name: instruction
    dtype: string
  - name: answers
    sequence: string
  splits:
  - name: train
    num_bytes: 25901752.319920518
    num_examples: 1922
  - name: test
    num_bytes: 1347645.8022851464
    num_examples: 100
  download_size: 21627134
  dataset_size: 27249398.122205663
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
  - split: test
    path: data/test-*
---